BSSPAR S11 April 2005 Questions and answers Q: Is there any counter about RXLev in UL in Idle Mode? A: There is any counter to measure the RXLEV in idle mode. In idle the MS does not transmit anything in UL (except some location area update on the SDCCH when changing location areas or triggered by the periodic update) so the BS can not measure anything. In the DL the MS measures the RXLEVs in order to know in which cell to camp but it does not report it to the network. The only measurement I know in idle is the interference measure in UL by the BS on the empty timeslots Q: More on FACCH Call Setup A: The FACCH physical channel is the same than the TCH, the only difference is the use. When the network needs to send signaling (i.e. FACCH call set up or HO signaling) it steals TCH frames using the stealing flag bit to indicate that is sending signaling instead of speech. Q: Why not using always variable step size in Power Control algorithm instead of fixed step size? A: The answer is that the philosophy of the PC algorithm is to smoothly adapt to power fluctuations when the power difference is not that big. By using small steps, it approaches to current value (although it might need several steps). Using always variable sizes to reach at each moment the right step size could cause many power oscillations. At any rate, note that when there is a significant power difference, a variable step size is used. Q: Is AMH used only in IUO? A: (From NED 11) Advanced Multilayer Handling (AMH) concept consists of three different features, which all are related to network load. The AMH features are: • BSC-initiated Traffic Reason Handover. For more information, see RF Power Control and Handover Algorithm . • IUO load control. For more information, see Intelligent Underlay-Overlay . • Multilayer load control. For more information, see Dual Band Network Operation. The AMH provides the operator both improved quality and capacity by distributing the traffic evenly over the network. It is useful especially in multiband (GSM 900/GSM 1800), microcellular or multilayer (Intelligent Underlay-Overlay) networks. Q: Why Directed Retry is not enabled automatically in the network? What is the disadvantage (if there is one)? A: In DR, the target cell is not the best one from a RXLev point of view; so if we enable DR, we take the risk of establishing a call in a TCH that is not optimised for that MS. By the way, one can object that it is always better to establish a call in non-optimised condition that not to have it established.
The practical reason is that actually Nokia wants to take out money from selling that feature. Q: Problem found in the network: There are a lot of SDCCH drops at call establishment with the reason "SDCCH Abis fail call"; that counter is very high (too high, probably there is a problem in the counter). They opened a case with Nokia, and Nokia says that it may be something related to RACH, but ghost RACH counter is 0. A: Nokia's right. The counter is pegging ghost SDCCH seizures as Abis failures. A right estimation of Abis Failures would be SDCCH Abis Fail Call - t3101_expired Q: In IUO, how is the interference measured? On the BCCH channel? If it is so, how to measure the interference on the hopping TRXs that do not carry BCCH? There must a method to do that because IUO allows intracell handover between different TRXs belonging to the same frequency group or not. A: IUO interference estimate is as follows : C/I (n) = Serving cell RXLEV - Adj cell (n) RXLEV + Serving TCH PC power reduction. So it is the DL C/I that user would face if the neighboring cell in question would use co-channel and users would be on full power. The interfering cells are defined separately for each super layer TRX. If TRX has different frequency (non-hopping case) it can have different interfering cells. In that case also the FRT would be defined to be different in these TRXs. In general IUO/IFH is nightmare to manage and gains can be small or non-existent especially if PC is in use, so not very highly recommended apart from some special cases perhaps. Q: Is GPRS active in the Super Reuse Layer in IUO? A: GPRS is not allowed on super layer TRXs. BSC does not allow it even if you try Q: When is Inverse ICE (COCA) adviced to be used? In what situation? A: This can be useful if: - BCCH has tight frequency reuse and is interfered on cell edges (in the US: narrow band network in which I need a minimum number of frequency reserved for hopping; e.g., having only 25 frequency available, 15 can be assigned to BCCH and 10 to hopping, but the latter assignment means that I cannot have more than 10 hopping TRXs in one site in both RF and BB hopping, so I need more frequencies assigned to hopping: 15 to hopping and 10 left for BCCH. But in the latter assignment BCCH has higher interference). - Network is coverage limited so that FH gains help in cell edge (note that these FH gains would only be visible in FER and not in RXQUAL statistics, also DCR could improve and it could be possible that improved DCR causes calls in marginal situations to stay on longer generating more bad quality samples and RXQUAL could even degrade) Q: Considering a network with only macrocells, what is the mechanism to redirect some traffic (handover) when a cell is overloaded from macrocell to macrocell? A: One is Direct Retry in case of congestion. Besides there are BSC traffic reason HO (AMH) and there also MSC controlled traffic reason HO. Nokia doesn't recommend the last two at the same time.